Towards a Semantic Web. A comparison between RDA and REICAT descriptive solutions

0.52 MB·Italian

by  Simone Forassiepi

Towards a Semantic Web.
A comparison between RDA and
REICAT descriptive solutions
Simone Forassiepi
1 Introduction
Making data fully interoperable in the semantic web is the greatest
challenge that libraries are currently facing. The transition from
rigidly structured record on proprietary formats and their diffusion
only in the library field to linked open data, interconnected with the
rest of the web, represents a radical transformation in the
organization of cataloguing information.
In order to do this, libraries must build their structured data on
logical entities clearly defined and globally shared, in order to
facilitate the construction of semantic ontologies which could be
used even in areas outside the original domains.
The creation of an ontological map of the bibliographic world is the
first step to take to create a community library that could interact
with the web through a mutual exchange of data. In order to begin
an alignment process among logical entities produced by RDA
drafters (RDA 2010) and those at the core of the new consolidated
edition of the ISBD (International Federation Of Library Associations
and Institutions 2011), in November 2011, a meeting was held in
Glasgow for the harmonization of data among the ISBD Review
Group, the ISSN Network and RDA Joint Steering Committee. This
JLIS.it Vol. 6, n. 1 (January 2015)
DOI: 10.4403/jlis.it-9963
S.Forassiepi, Towards a Semantic Web.
meeting produced a table of comparison1 whose main purpose is to
be the first step towards full interoperability between the two texts.
Starting from the structure of the ISBD areas, in this paper the
priority is to show how the logical entities described in REICAT,
chapter four (Commissione permanente per la revisione delle regole
italiane di catalogazione and Istituto centrale per il catalogo unico
delle biblioteche italiane e per le informazioni bibliografiche 2009),
could be collocated in this alignment process. In the following pages,
we will compare the textual architectures, the basic logical entities of
RDA, the ISBD consolidated edition and REICAT with one another,
analysing the similarities and the differences, to try to understand
whether from the new Italian cataloguing rules could emerge a
cataloguing structure fully operational in a semantic environment.
2 Area 0
The ISBD area 0 is devoted to the GMD, General Material
Designation. This new area identifies the described resource's
product group, defining a documentary macro-category for the
identification of the typology of the resource. These data, which has
been present in the ISBD from 1977, have entered in the ISBD schema
for the first time with the ISBD Consolidated edition. After a
temporary collocation in area 1, the GMD have been systematized in
the new area 0.
The area is divided into two sections: content form and media type. The
former defines the resource expressive form, «form or forms in
which the content of a resource is expressed » (ISBD. 0.1), while the
latter, devoted to the media type, is for entering «the type or types of
carrier used to convey the content of the resource» (ISBD. 0.2). For
both sections there is a list of controlled terms to be used.

1 Alignment of the ISBD element set with RDA element set – RDA, Appendix D.1.
URL: http://www.rda--jsc.org/2011jscisbdissnoutcomes.html.
JLIS.it. Vol. 6, n. 1 (January 2015). Art. #9963 p. 34
JLIS.it. Vol. 6, n. 1 (January 2015)
The need to put in a "zero" position the instructions about the
material form of the resource arises as a result of an ever-increasing
amount of non-textual resources appearing in the collections
described by the ISBD. In a semantic environment, where the
metadata produced by libraries happen to be linked to data coming
from different domains, the designation of the material becomes a
key element for the resources identification.
Regarding content form and media type, the RDA Joint Steering
Committee decided to separate the information in two different
portions of the text,2 devoting chapter 3 (Describing carriers) to the
media type, and the paragraph 6, 9 (Content type) to the content form.
This relocation implies a deep conceptual modification. The content
form analysis is not any longer connected to manifestation but it is
dealt within the chapter devoted to works and expressions. The
reason for this relocation is the willingness to build a catalogue
structure based on the four different typologies of the resource
inspired by FRBR (IFLA Study Group on the Functional
Requirements for Bibliographic Records et al. 1998).
The RDA architecture aims at increasing interoperability with data
produced in domains external to the libraries. In this way the
cataloguer analysis is extended to resources that are not always
based on the typical bibliographic quadripartite structure. Finally,
we want to emphasize that RDA drafters "explode" the record
structure typically identified by the ISBD, by changing the sequence
of the elements. At the base of this choice there is a focus change, the
attention of the text is no longer focused on the information
structure, but on the individual data. In this perspective, the
sequence of the areas, a key element for the ISBD world, loses his

2 RDA has 37 chapters divided into 10 sections. The first two are devoted respectively
to the Recording attributes of manifestation and item and to the Recording attributes of
works and expressions.
JLIS.it. Vol. 6, n. 1 (January 2015). Art. #9963 p. 35
S.Forassiepi, Towards a Semantic Web.
centrality and the visual organization of data becomes a simple
layout choice made by the user.
In the Italian rules, the GMD does not find an autonomous space,
unlike the great relevance reserved to it in the other two texts, and it
has been relegated to search filter: «the general material designation3
is not covered by these rules as part of the bibliographic description.
It is considered more appropriate to record it separately, usually in
coded form, in order to display it according to the procedure and in
the most convenient position, as well as possible filter for the search»
(REICAT, 4.1.0.1, footnote 1). This choice meets the specific practice
of the Italian libraries, still strongly focused on the cataloguing of the
print resources, but it is lacking in the theoretical aspect and in the
prospect of a cataloguing functional to new semantic platforms4.
3 Area 1
Compared to the past, in Area 1 there is a radical lexical change
made by the ISBD. The standard changes its focus, with regard to the
statement of responsibility from the term author to the act of
creation, «a statement of responsibility consists of one or more
names, phrases, or groups of characters relating to the identification
and/or function of any persons or corporate bodies responsible for or
contributing to the creation or realisation of the intellectual or artistic
content of a work contained in the resource described» (ISBD., 1.4).
This definition leads to the use of the term creator instead of author, a
term rarefied in a list of figures (Group and Committee 2011, chap.
1.4.2) and it is unchanged in RDA. Regarding REICAT, the term
creator does not appear anywhere in the text, but the meaning given

3 As for the adjective that qualifies the GMD, departing from the old translations and
from the REICAT choices for the Italian translation of the ISBD consolidated edition is
selected to use the term generale (general).
4 «The rules are based on the needs of a detailed, rigorous and uniform cataloging that
characterize the catalogs of library systems or cooperation networks and those of large
libraries» (REICAT 0.1.4 A).
JLIS.it. Vol. 6, n. 1 (January 2015). Art. #9963 p. 36
JLIS.it. Vol. 6, n. 1 (January 2015)
to statement of responsibility is very close, «by statement of
responsibility we mean a name or an expression that indicates
persons or entities that have a relationship of responsibility with the
published works or their expressions, or function they perform»
(REICAT, 4.1.3.0). The slip from the concept of author to that of
creator has resulted an extension of the concept of responsibility5 in the
three texts, especially in the field of audio-visual resources that often
see the efforts of several individuals. The creator concept results,
compared with that of author, closely connected to work and
expression level rather than manifestation. Furthermore if the
changes appear minimal in a descriptive level, the same cannot be
said on the conceptual. The creator figure is connected to all those
personalities who play a creative role in the definition of a resource,
and not anymore at the single author. The choice of REICAT to not
ever refer to the concept of creation is muddled and disconnected
from the international context. Also clearly demonstrates the lack of
propensity to create ontologically well-defined elements, the first
step towards a construction of a semantic interface.
Regarding the textual architecture, RDA has chosen to divide the
area into two parts 1: Title (2.3) and Statement of Responsibility
(2.4).6 This decision once again demonstrates the perspective shift
made by RDA with the change from a focus centred to the record to
a data centrality. This split is based on the need to divide the
elements in order to better identify the individual entity to be
marked. Title and Author have been included in the same area since
the first edition of the ISBD as they are considered the most
significant elements for the identification of a resource. RDA breaks

5 The Italian choice is complex, because the use of the entity creator would have been a
logical simplification compared to the use of a statement of responsibility. It also appears
the lack of interest towards the creation of ontologically-defined terms usable in a
semantic structure.
6 The second chapter of the RDA, dedicated to the description of the Manifestation and
Item, broadly reflects the structure of ISBD areas and when this does not occur implies
a conceptual shift.
JLIS.it. Vol. 6, n. 1 (January 2015). Art. #9963 p. 37
S.Forassiepi, Towards a Semantic Web.
this habit. The use of an XML metalanguage provides that each data
should be "atomized" in order to be well defined and used in a
semantic architecture.
Area 2
This ISBD section is strictly connected to hardcopy resources. For
this reason, the differences among RDA, REICAT and the ISBD are
minimal. This area is based mainly on information available in
books. Therefore, this area is less subject to a data exchange with
areas external to the libraries. As a consequence, there is no deep
conceptual difference among the three codes, which do not show the
usual differences emerging from the greater or lesser openness
towards resources external to libraries.
Area 3
The third ISBD area is restricted to cartographic resources, notated
music and serials. In the consolidated edition, the fourth resource,
that was traditionally part of the Area 3, the electronic resources, has
been placed elsewhere. Area 3 examines the mode of presentation of
the characteristic data, observing mathematical data for cartographic
resources (scale, projection, coordinates and equinox), musical
format statement for notated music (as score, musical arrangement,
etc…) and numbering related to serials.
REICAT drafters follow closely, in this portion of the text, the ISBD
structure, with the only difference represented by the fact that firstly
REICAT deals with notated music and secondly with cartographic
resources.
Regarding RDA, even in this case, the text drafters decided to “blast”
the area, as they did for area 0. The serials are being discussed in the
second chapter, paragraph 2.6 Numbering of Serials, while data
related to cartographic resources and to notated music are collocated
in the portion of the text devoted to works and expressions (RDA 7.4
JLIS.it. Vol. 6, n. 1 (January 2015). Art. #9963 p. 38
JLIS.it. Vol. 6, n. 1 (January 2015)
Coordinates of cartographic content, 7.5 Equinox, 7.6 Epoch, 7.20
Format of notated music).
As for the analysis of the elements present in this area, a deep
difference can be noted between the ISBD and RDA. There are two
completely different views of the serials in the two texts. According
to RDA drafters, the serials are split and identified in their basic
elements (the singles numbers), while, in the ISBD (3.3.2 and 3.3.3)
and in REICAT (4.3.C.1.0), the series are identified and reported by
the first number of the collection and by the last one. In this way, all
the collection is perceived as a unique entity. From this structure
derive two different cataloguing processes that are considered as one
of the most peculiar differences between RDA and the ISBD, and
likewise between RDA and REICAT.
This different treatment once again emphasizes the willingness of
data atomization carried out by RDA drafters, in order to make the
singles numbers individually indexed. On the contrary, the creation
of a range of numbers, referred to a series, makes it impossible to
search internally for individual numbers.
Area 4
The forth area aims at covering «all types of publication, production,
distribution, issuing and release activities connected with resources»
(ISBD. 4. Introductory note), and data related to manufacturing.
Regarding this area, it is to be pointed out an important fracture
between the ISBD and REICAT on one hand, and RDA on the other.
If in the two former texts the concepts of publication, production, print
and manufacture are gathered in one section of the text, in RDA the
drafters decided to assign a paragraph to each figure. Accordingly,
RDA emphasizes the difference among the figures of publisher,
producer, distributor and printer.
Such a specific subdivision of the roles within the production process
is oriented towards the elements atomization, which in this way they
JLIS.it. Vol. 6, n. 1 (January 2015). Art. #9963 p. 39
S.Forassiepi, Towards a Semantic Web.
can undergo a process of autonomous tagging. Each element is
thereby identified as a "quid" in its own right, in which the
punctuation and the quotation order (ISBD cornerstone) loses
importance, becoming simple elements of a layout changeable at
user discretion. The catalogue is focused, once again, on the data and
not on the record. It is the single informative element that will be
analysed and structured, not the record structure.
The RDA main goal is to establish a unique set of entities
individually marked with a specific URI. It is no longer the area
context to define the role of a term but it is the same term that
defines itself through its own autonomous tag. This paradigm shift is
central to the creation of an ontological grammar of the cataloguing
world, the first step towards the loss of the catalogue form in favour
of a data cloud structure.
Focusing on two paradigmatic choices made by RDA and REICAT
drafters, with regard to the area 4, can help us to understand the
different perspectives in the information organization offered by the
two codes. RDA distinguish themselves among the examined texts
since they include the concept of parallel place of manufacture, among
the analysed entities. This shows a willingness to cover all the
possible options with the purpose of a complete tagging. On the
other hand, REICAT are the only text in which the place of printing
and the name of the printer are dealt with together in a single chapter.
Although the same ISBD cataloguing choices are made, from the
Italian text emerges a more interesting perspective to identify the
resource properties and to build a record as complete as possible. On
the contrary, RDA loses interest in the record construction and it
turns to the unambiguous data marking.
The second element used for showing the differences between RDA
and REICAT is found in the following statement of the Italian rules
«as editor means the responsible for the publication of the issue. It
may be a commercial editor, a public or private entity of any nature,
a person, or more institutions or persons presented as responsible for
JLIS.it. Vol. 6, n. 1 (January 2015). Art. #9963 p. 40
JLIS.it. Vol. 6, n. 1 (January 2015)
the publication, regardless of the functions actually performed»
(REICAT 4.4.2.0). At first sight, this definition given by REICAT is
very accurate and not too far from that provided by RDA (2.7.4). The
distance between the two texts emerges when in REICAT is pointed
out that in the entity of the publisher «are included distributors,
booksellers and other figures that appear in relationship with the
purchasers, the diffusions or published commercialization» (REICAT
4.4.2.0). As a consequence, the diversity of the roles in the process of
production and distribution of a resource is lost, while in RDA
autonomous chapters are found for each entity. On the other hand,
the roles are defined better in the ISBD than in REICAT, but they are
contained in a single entity: «The name of the person or corporate
body appearing on the resource that effects respectively the
publication, production and/or distribution or release activities for
the resource» (ISBD. 4.2). Between the ISBD and REICAT on the one
hand and RDA on the other emerges a radical split, as Resource
Description and Access clearly distinguishes the different figures and
creates individual definitions for producers (RDA 2.7.4), publishers
(RDA 2.8.4) and distributor (RDA 2.9.4). The RDA main goal is once
again the identification of a single information, in the perspective of
the creation of a linked data network, and that is how it clearly
distinguish itself from the other two texts.
Area 5
The area 5 is the ISBD section based on the material description and
includes «the extent, other physical details, the dimensions, and the
accompanying material statement» (ISBD. 5. Introductory note). Due
to the variety of analysable resources, this area has a great
importance and clarifies the codes perspectives as far as the future
interests are concerned.
In the section of the text devoted to the description of the material,
the ISBD and REICAT have overlapping structures and even the
discussed entities are ontologically very close. The first part of the
JLIS.it. Vol. 6, n. 1 (January 2015). Art. #9963 p. 41
S.Forassiepi, Towards a Semantic Web.
area is devoted to the specific material designation (SMD) and to the
extent. Then there are the recommendations about other physical
details and dimensions. Finally, both texts end by dealing with
accompanying material statement.
RDA is detached from this pattern, devoting a separate chapter to
the material description, the third one, Describing Carriers. It was
decided by the drafters not to include the material description in the
paragraphs sequence of the second chapter devoted to the
identification of Manifestation and Item and abandon the parallelism
with the ISBD areas just to express the importance of the element.7
The use of the rules on platforms used outside the library domains
makes the material description, along with the GMD, an element of
primary importance. RDA drafters, being more inclined to open
towards new fields, build a much more precise text in the
specification of the material description, so that in the text appears a
division among base material, applied material and mount material.
An element of particular importance, to show the three texts
different perspectives in the discussion about the elements of the
fifth area is the behaviour of the rules towards the fixed and moving
images. The ISBD, debating the issue of colour, points out a single
choice between colour and black and white. In REICAT the choice is
extended to another “colour” possibility, the sepia, and with the
chance to mention the system of colour reproduction, e.g.
Technicolor (REICAT 4.5.4.5 B). RDA proves to be, once again, like
the most adjustable rules, as they speak explicitly of «presence of
colours, tones, etc…» and from the examples it draws a complete
freedom to define the colour treatment.
RDA remarkable peculiarity to work with a great amount of
resources is also evident at paragraph 3.16.2, where, as far as the
recording method is concerned, the rules refer clearly to digital, while

7 Chapter 3 is in the first section of the text.
JLIS.it. Vol. 6, n. 1 (January 2015). Art. #9963 p. 42